TWI761863B - Traffic condition detection method - Google Patents

Abstract

A traffic condition method, comprises: obtaining a plurality of traffic parameters associated with a monitoring area, and obtaining a normal parameter range based on the traffic parameters, wherein at least half of the traffic parameters fall within the normal parameter range, and performing a monitoring procedure on the monitoring area. The monitoring procedure comprises: determining whether an instant traffic parameter falls within the normal parameter range, and outputting a traffic abnormality notification when the instant traffic parameter does not fall within the normal parameter range.

Description

Traffic situation detection method

The present invention relates to a traffic condition detection method, in particular to a traffic condition detection method that can use different monitoring standards for different monitoring areas.

As the traffic field becomes more and more complex, in order to monitor each traffic field more efficiently, manufacturers develop various systems for monitoring the traffic field one by one. For example, the existing monitoring method includes first selecting a region of interest (ROI) on the monitoring screen, and then tracking objects such as vehicles and locomotives in the region of interest to output the objects in the region of interest. Traffic parameters in the area (for example, vehicle speed, direction of travel, traffic flow, etc.), the monitoring personnel of the traffic monitoring center can judge whether there is an abnormal traffic situation based on the traffic parameters.

However, according to the existing monitoring method, it is still necessary to rely on the monitoring personnel to interpret the traffic parameters to determine whether there is an abnormal traffic situation. In addition, although the standard values of basic traffic parameters can be set so that the system can interpret simpler traffic parameters based on the standard values, the standard values of different road sections are also different, so the standard values still need to be set manually, which not only makes In the initial stage of erection and subsequent actual monitoring process, the monitoring system still needs manpower to complete the above work, and in the initial stage of erection and monitoring process, the monitoring results may be inaccurate due to human error.

In view of the above, the present invention provides a traffic condition detection method to meet the above requirements.

A traffic condition detection method according to an embodiment of the present invention includes: obtaining a plurality of traffic parameters associated with a monitoring area, and obtaining a normal parameter range based on the traffic parameters, wherein at least half of the traffic parameters fall within within the normal parameter range; and executing a monitoring program for the monitoring area, wherein the monitoring program includes: judging whether a real-time traffic parameter in the monitoring area falls within the normal parameter range; and when the real-time traffic parameter does not fall within the normal parameter range When within the normal parameter range, output a traffic abnormality notification associated with the monitoring area.

To sum up, according to the traffic condition detection method shown in one or more embodiments of the present invention, suitable normal parameter ranges can be established for different monitoring areas to apply different monitoring standards for different monitoring areas, and Different monitoring criteria are applied at different times (eg, peak hours, off-peak hours). When the environment of the monitoring area changes, the normal parameter range can be updated with appropriate traffic parameters, so as to maintain the normal parameter range in the current state of the monitoring area. In addition, when the traffic condition sensor is initially set up, the monitoring program can be executed for the monitoring area. In addition, when it is determined that there is an abnormal condition in the monitoring area, the monitoring personnel can be notified in time to implement contingency measures, and at the same time, inaccurate monitoring results can be avoided by artificially determining whether the traffic parameters are abnormal.

The above description of the present disclosure and the following description of the embodiments are used to demonstrate and explain the spirit and principle of the present invention, and provide further explanation of the scope of the patent application of the present invention.

The detailed features and advantages of the present invention are described in detail below in the embodiments, and the content is sufficient to enable any person skilled in the relevant art to understand the technical content of the present invention and implement it accordingly, and according to the content disclosed in this specification, the scope of the patent application and the drawings , any person skilled in the related art can easily understand the related objects and advantages of the present invention. The following examples further illustrate the viewpoints of the present invention in detail, but do not limit the scope of the present invention in any viewpoint.

The traffic condition detection method shown in the present invention is executed by, for example, a central processing unit, a server, and other devices with computing capability of a traffic monitoring center. For ease of understanding, the following description will take the server as the device for executing the traffic condition detection method, but the present invention is not limited thereto.

Please refer to FIG. 1 . FIG. 1 is a flowchart of a traffic condition detection method according to an embodiment of the present invention.

Step S10: Acquire a plurality of traffic parameters associated with the monitoring area.

The traffic parameters are obtained, for example, based on sensing data measured by a traffic condition sensor, wherein the traffic condition sensor is, for example, a camera or a speedometer. For example, when the traffic condition sensor is a camera, the monitoring area can be the entire area within the shooting range of the camera, or a part of the area within the shooting range of the camera (such as one of the roads, parking spaces, etc.) , a sidewalk or a part of a road, etc.), so the sensing data measured by the traffic condition sensor is, for example, an image captured by a camera.

The server is connected to the traffic condition sensor in communication to obtain sensing data such as images. After the server obtains the images, it will perform image recognition on the objects located in the monitoring area in the images, and track the recognized objects. In order to determine the moving speed of the object, the moving direction, the stay time in the monitoring area, the ratio of the number of objects in the monitoring area to the area of the monitoring area, and the number of objects passing through the monitoring area in a period of time, etc., and the said These parameters can be used as the traffic parameters.

Similarly, when the traffic condition sensor is a speedometer, the moving speed of the object sensed by the speedometer can also be used as the traffic parameter, which is not limited in the present invention.

Step S20: Obtain the normal parameter range based on the traffic parameters.

After obtaining the traffic parameters, the server can integrate at least half of the traffic parameters into a parameter range, and use this parameter range as a normal parameter range. In other words, at least half of the traffic parameters fall within the normal parameter range, wherein the detailed implementation of obtaining the normal parameter range based on the traffic parameters will be described in detail in the embodiments of FIGS. 3A and 4A .

Step S30: Execute the monitoring program on the monitoring area.

After the server obtains the normal parameter range, it can conduct real-time monitoring of the objects in the monitoring area by judging whether the real-time traffic parameters of one or more objects in the corresponding monitoring area fall within the normal parameter range, and based on the Corresponding measures are performed, wherein the monitoring procedure will be described in detail in the embodiment of FIG. 5 .

Please refer to FIG. 2 first. FIG. 2 is a flowchart of a method for obtaining a plurality of traffic parameters according to an embodiment of the present invention. That is, the method for obtaining a plurality of traffic parameters in step S10 of FIG. 1 can be obtained by the method shown in FIG. 2 in addition to the traffic condition sensor as described above. Monitoring can be started at the early stage of erection, so as to save the time of collecting multiple pieces of sensing data and then analyzing them.

Step S101: Calculate the degree of difference between the first traffic parameter associated with the first traffic object in the first monitoring area and the second traffic parameter associated with the second traffic object in the second monitoring area.

It should be noted that the first monitoring area and the second monitoring area are different monitoring areas. The monitoring area for which the monitoring procedure is performed based on the traffic parameters sensed from the second monitoring area. The type of the first traffic object is similar to the type of the second traffic object. For example, the first traffic object and the second traffic object may be both a car, a locomotive, a pedestrian, or a vehicle with wheels. That is, the first traffic object and the second traffic object are preferably the same or similar types of traffic objects, but the present invention does not limit the types of the first traffic object and the second traffic object.

To be more specific, for the server to determine the degree of difference between the traffic conditions of the first monitoring area and the second monitoring area, the implementation can be, for example, that the server obtains the first traffic of the first traffic object in the first monitoring area. parameters, and the second traffic parameter of the second traffic object in the second monitoring area, the degree of difference between the first traffic parameter and the second traffic parameter is calculated as the aforementioned degree of difference.

For example, the first traffic object and the second traffic object are, for example, cars; the first traffic parameter is, for example, a vehicle speed of 80km/h, and the second traffic parameter is, for example, a speed of 85km/h, then the server can convert the 80km/h The first traffic parameter of 85km/h and the second traffic parameter of 85km/h are subtracted to obtain the difference degree of 5km/h, wherein the difference degree obtained by subtraction is only an example, and the difference degree can also be the subtraction value of the two traffic parameters It is obtained by dividing by the first or second traffic parameter, or even obtaining the degree of difference by a more complicated calculation formula. The present invention does not limit the form of the degree of difference.

It should be noted that the aforementioned vehicle speed takes a single first traffic parameter and a single second traffic parameter as an example, and the number of the second traffic parameters obtained by the server is preferably greater than the number of the first traffic parameters. Therefore, the degree of difference is preferably calculated based on the set of first traffic parameters and the set of second traffic parameters (eg, the average value of multiple first traffic parameters and the average value of multiple second traffic parameters).

Step S102: Determine whether the degree of difference is not greater than a threshold.

After calculating the difference, the server can determine whether the difference between the first traffic parameter in the first monitoring area and the second traffic parameter in the second monitoring area is not greater than a threshold, where the threshold is used to represent the two traffics The acceptable degree of difference between parameters.

Step S103: Use the first traffic parameters as the traffic parameters.

When the difference is greater than the threshold, it means that the difference between the traffic conditions of the first monitoring area and the second monitoring area is too high (the similarity is low), so the server does not refer to the traffic parameters of the second monitoring area, but sets the The first traffic parameters sensed by the traffic condition sensor in the first monitoring area are used as the traffic parameters, and step S20 shown in FIG. 1 is performed after the traffic parameters are obtained.

Step S104: Use a plurality of second traffic parameters corresponding to the second monitoring area as the traffic parameters.

In other words, when it is judged that the degree of difference is not greater than the threshold, it means that the traffic conditions in the second monitoring area are similar to the traffic conditions in the first monitoring area. The controller can take the second traffic parameters of the second monitoring area as the traffic parameters, and execute step S20 shown in FIG. 1 after obtaining the traffic parameters.

To sum up, the method for obtaining a plurality of traffic parameters shown in FIG. 2 can be used for the first monitoring area in the early stage of the erection of the traffic condition sensor, so although the number of obtained first traffic parameters is still insufficient for the subsequent steps S20 obtains the corresponding normal parameter range, or the number of the first traffic parameters is still not enough to obtain the normal parameter range with higher reference degree, the second traffic parameter of the second monitoring area with similar traffic conditions can still be used as Establish traffic parameters within the normal parameter range of the first monitoring area, so that the server can execute the monitoring program on the first monitoring area as soon as possible, thereby reducing the time from collecting enough first traffic parameters to obtaining the normal parameter range corresponding to the first monitoring area time. In addition, by first comparing the difference between the first traffic parameter and the second traffic parameter, it is possible to effectively reduce the execution time when applying the second traffic parameter or the normal parameter range of the second monitoring area to the first monitoring area. Errors while monitoring the program.

Please refer to FIG. 3A first. FIG. 3A is a flowchart of a method for obtaining a normal parameter range according to an embodiment of the present invention. That is, FIG. 3A is one of the ways to realize step S20 of FIG. 1 .

Step S201: Screening out a plurality of time period traffic parameters associated with the time period separation parameter from the traffic parameters according to the time period separation parameter.

The time period separation parameter is a parameter used to separate different time periods, so as to filter out traffic parameters corresponding to different time periods. For example, if the time interval parameter is two time points, such as 8:00 am and 10:00 am, the server can filter out the traffic parameters according to the time interval parameter to select the time between 8:00 am and 10:00 am. The obtained traffic parameters, and the traffic parameters falling between 8:00 and 10:00 in the morning are used as the traffic parameters for a plurality of time periods.

Step S202: Form a normal distribution model with the traffic parameters of these time periods.

The server can divide the period between 8:00 am and 10:00 am into multiple sub-periods, and create a histogram corresponding to these sub-periods with the traffic parameters of these periods, and then connect the blocks of each square in the histogram with a straight line or a curve. The midpoint of the top edge is used to form a normal distribution model, which is a model composed of traffic parameters during the period from 8:00 to 10:00 in the morning.

Step S203: Take the confidence interval of the normal distribution model as the normal parameter range.

Please also refer to FIG. 3B . FIG. 3B is an example diagram according to the normal distribution model ND shown in FIG. 3A . The vertical axis is, for example, the cumulative number of objects of the time period traffic parameters of each interval. For example, when the traffic parameter is vehicle speed, the horizontal axis of the normal distribution model ND is, for example, the distribution range of vehicle speeds from 8:00 am to 10:00 am (for example, 15 kilometers per hour to 40 kilometers per hour); and the normal distribution model ND The vertical axis is, for example, the cumulative number of vehicles whose speed falls within each speed interval (for example, the cumulative number of vehicles whose speed falls between 15 kilometers per hour and 20 kilometers per hour, and the cumulative number of vehicles whose speed falls between 20 kilometers and 25 kilometers per hour). quantity, etc., etc.).

After the normal distribution model ND as shown in FIG. 3B is established, the server can use the confidence intervals CI1 to CI3 of the normal distribution model ND as the normal parameter range, wherein the confidence intervals CI1 to CI3 are, for example, traffic in these periods of time The average AVG of the parameters plus or minus one or more standard deviation S speed interval (for the convenience of understanding, the average AVG is the value without adding or subtracting the standard deviation S, so the average AVG in FIG. 3B is represented by the value "0", the present invention does not mean The actual value of AVG is limited). In other words, the normal parameter ranges are, for example, the 68% confidence interval CI1 of the mean AVG plus or minus three standard deviations (+3S and -3S), and the 95% of the mean AVG plus or minus two standard deviations (+2S and -2S). The confidence interval CI2 or the 99.7% confidence interval CI3 of the mean AVG plus or minus one standard deviation (+1S and -1S), the present invention does not limit the value of the confidence interval.

Please refer to FIG. 4A , which is a flowchart of a method for obtaining a normal parameter range according to another embodiment of the present invention. That is, FIG. 4A is one of the ways to realize step S20 of FIG. 1 .

Step S201': According to the distribution state of the traffic parameters corresponding to the time parameters, filter out a plurality of time period traffic parameters associated with the time period from the traffic parameters.

In other words, while acquiring the traffic parameters, the server also records time parameters corresponding to each of the traffic parameters (ie, the time when each of the traffic parameters was sensed). Therefore, the server can first establish the distribution state of the traffic parameters corresponding to the time parameters.

Please refer to FIG. 4B together. FIG. 4B is an exemplary diagram according to the distribution state shown in FIG. 4A . Taking Figure 4B as an example, the unit of the time parameter is "time", and the unit of the traffic parameter is the speed of "km/h", that is, the range of the horizontal axis of the distribution state is, for example, 6:00 am to 20:00 pm; The range of the axle is, for example, from 25 kilometers per hour to 55 kilometers per hour.

According to the distribution state shown in Fig. 4B, the server can determine that the traffic conditions from 8:00 am to 10 am is different from the traffic conditions from 10:00 am to 17:00 pm by the slope changes at each time or time period. A plurality of time-period traffic parameters corresponding to the time period T1 from 8:00 am to 10:00 am are further screened, and a plurality of time-period traffic parameters corresponding to the time period T2 from 10:00 am to 17:00 pm are filtered out.

Please go back to FIG. 4A and refer to step S202': forming a normal distribution model with the traffic parameters of these time periods; and step S203': using the confidence interval of the normal distribution model as the normal parameter range.

After filtering out the time period traffic parameters corresponding to different time periods, the server can then execute steps S202' and S203' to obtain the normal parameter range, wherein the implementation of steps S202' and S203' is the same as steps S202 and S203 in FIG. 3A , Therefore, the implementation manners of steps S202' and S203' are not repeated here.

In addition, similar to the embodiment of FIG. 2 , when the traffic condition sensor in the first monitoring area is in the early stage of erection and the number of the first traffic parameters is still insufficient to establish a normal parameter range with a reference degree, the server can also Steps S101 to S102 shown in FIG. 2 are performed, and when it is determined that the degree of difference is not greater than the threshold, the normal parameter range of the second monitoring area is directly used as the normal parameter range of the first monitoring area.

Please refer to FIG. 5A first. FIG. 5A is a flowchart of a monitoring program according to an embodiment of the present invention. That is, FIG. 5A is one of the ways to realize step S30 of FIG. 1 .

Step S301: Determine whether the real-time traffic parameters in the monitoring area fall within the normal parameter range.

After the server obtains the real-time traffic parameters by sensing the objects in the monitoring area in real time through the traffic condition sensor, the server can then determine whether the real-time traffic parameters fall within the normal parameter range. For example, the monitoring area is, for example, in front of a traffic light at an intersection, the normal parameter range is, for example, the average time that the vehicle stays for 40 seconds to 60 seconds when the traffic light is red, and the real-time traffic parameter is when the traffic light is red. The dwell time of the object being sensed before the traffic light when the number is counted.

Step S303: Update the normal parameter range with real-time traffic parameters.

When the server determines that the real-time traffic parameters fall within the normal parameter range, it means that the state of the object in the monitoring area conforms to the normal state of the monitoring area, and the server can update the normal parameter range with real-time traffic parameters.

Taking the above dwell time before the traffic light as an example, when the server determines in step S301 that the real-time traffic parameter of 55 seconds falls within the normal parameter range of 40 seconds to 60 seconds, the server can further update the normal parameter in 55 seconds range, the median and/or average value of the normal parameter range will increase slightly, and when the confidence interval ratio of the normal parameter range does not change, the 40-second boundary value of the normal parameter range will also be updated Then increase, so as to keep the normal parameter range continuously in line with the current traffic state of the monitoring area.

其中，

為偵測到監控區域中實際存在車輛的時間的總和；

為監控區域中假定的車輛的佔有率，且

可以為如圖2的實施例所示，依據其他相似的監控區域的佔有率所假設得的佔有率。換言之，

為給定

後

成立的機率；

為

成立時觀察到

的機率（即在假定

的情況下

成立的機率）；

為監控區域中實際存在車輛的時間的總和(

)對應總偵測時間的值（亦即，若總偵測時間為

，則

即為

）；

為觀察到

前，

成立的機率（即在不考慮

的情況下，

成立的機率）。Additionally, the server may update the normal parameter distribution to predict its extent based on the real-time traffic parameters using Bayesian Inference. It should be noted that the Bayesian inference method can be used to train artificial intelligence networks (Artificial Neural Network, ANN), so that the server can predict more accurate values; and the Bayesian inference method described here is only an example, The server can also use other inference methods such as frequency inference, likelihood-based inference, Akaike information criterion, etc. and their branching methods to predict the normal parameter range. The normal parameter range is limited in the way. Taking the traffic parameters as the occupancy rate of vehicles in the monitoring area as an example, the formula for predicting the normal parameter range of the corresponding occupancy rate can be as follows:

in,

It is the sum of the time when the vehicle is actually detected in the monitoring area;

is the assumed occupancy rate of vehicles in the monitored area, and

As shown in the embodiment of FIG. 2 , the occupancy rate may be assumed based on the occupancy rate of other similar monitoring areas. In other words,

for a given

back

the probability of being established;

for

observed when established

the probability of (that is, assuming

in the case of

the probability of being established);

is the sum of the time when the vehicle actually exists in the monitoring area (

) corresponds to the value of the total detection time (that is, if the total detection time is

,but

that is

);

to observe

forward,

the probability of being established (that is, regardless of

in the case of,

probability of being established).

及

後，可以先據以計算出

、

，以及

，並接著如上述公式計算出

，而此述的

即為伺服器所預測得對應佔有率的常態參數範圍。因此，當伺服器計算出

後，伺服器即可進一步判斷

是否大於一預設值，並且當

大於該預設值時，則以

做為監控區域中的車輛的佔有率；若

小於該預設值時，則以另一假定

再次進行預測。Therefore, the server is getting

and

After that, we can first calculate

,

,as well as

, and then calculated as the above formula

, while the stated

It is the normal parameter range of the corresponding occupancy rate predicted by the server. Therefore, when the server calculates

After that, the server can further judge

is greater than a preset value, and when

When it is greater than the default value, the

As the occupancy rate of vehicles in the monitoring area; if

When it is less than the default value, another assumption is used

Make predictions again.

Step S305: Output a traffic abnormality notification associated with the monitoring area.

That is, when the server determines in step S301 that the real-time traffic parameter falls outside the normal parameter range, it means that the traffic state of the object in the monitoring area does not conform to the normal state of the monitoring area, and the server can output a traffic abnormality notification, wherein the traffic abnormality The notification preferably includes location information of the monitored area.

Taking the above dwell time before the traffic light as an example, when the server determines in step S301 that the real-time traffic parameter of 70 seconds is outside the normal parameter range of 40 seconds to 60 seconds, it means that there may be a vehicle breakdown or a car accident before the traffic light. Abnormal events, so the server can output traffic abnormal notification to the terminal device of the traffic monitoring center, wherein the traffic abnormal notification preferably includes the location information of the traffic light (monitoring area) and the staying time of the object in front of the traffic light (real-time traffic parameters), etc. To remind monitoring personnel that there may be abnormal traffic conditions to be excluded.

In addition, after judging that the real-time traffic parameters fall within the normal parameter range (step S301 ), and before updating the normal parameter range with the real-time traffic parameters (step S305 ), the server may further update the real-time traffic parameters that fall within the normal parameter range Multiply by a weight value greater than 1, and then update the normal parameter range by the real-time traffic parameter multiplied by the weight value, so that the updated normal parameter range is more in line with the current traffic conditions in the monitoring area.

Please refer to FIG. 5B . FIG. 5B is a flowchart of a monitoring program (step S30 ′) according to an embodiment of the present invention, wherein steps S301 ′, S303 ′ and S305 ′ in FIG. 5B are the same as steps S301 , S305 ′ in FIG. 5A . S303 and S305, so the similarities will not be repeated here. The difference between FIG. 5B and FIG. 5A is that when it is determined in step S301' that the real-time traffic parameters do not fall within the normal parameter range, the server executes step S304'.

Step S304': Determine whether the real-time traffic parameters fall within the buffer zone.

There can be buffers outside the normal parameter range, and the buffers are adjacent to the boundary value of the normal parameter range. In other words, when the server determines in step S301' that the real-time traffic parameters do not fall within the normal parameter range, the server may also execute step S304' to further determine whether the real-time traffic parameters fall within the buffer zone.

Taking the above dwell time in front of a traffic light as an example, the boundary values of the normal parameter range are 40 seconds and 60 seconds, and the buffer interval corresponding to the 40 second boundary value is, for example, 35 seconds to 40 seconds; is 60 seconds to 65 seconds, so when the server determines that the real-time traffic parameter of 63 seconds falls within the buffer of 60 seconds to 65 seconds, the server can perform step S303'; on the contrary, when the server determines that, for example, it is When the real-time traffic parameter of 30 seconds does not fall within the buffer of 35 seconds to 40 seconds, the server may execute step S305'. That is, when the server determines that the real-time traffic parameter does not fall within the normal parameter range, it can further determine whether the real-time traffic parameter falls within one of the two buffers.

In short, when judging that the real-time traffic parameters do not fall within the normal parameter range, the server may directly execute step S305 as shown in FIG. 5A to output a traffic abnormality notification, or may execute step S304 ′ as shown in FIG. 5B to first determine whether the real-time traffic parameters are not within the normal parameter range. If it falls within the buffer zone, step S303 ′ or S305 ′ is selected to be executed according to the judgment result of step S304 ′, so as to prevent the traffic monitoring center from continuously receiving traffic abnormality notifications and increasing the workload of monitoring personnel.

To sum up, according to the traffic condition detection method shown in one or more embodiments of the present invention, suitable normal parameter ranges can be established for different monitoring areas to apply different monitoring standards for different monitoring areas, and Different monitoring criteria are applied at different times (eg, peak hours, off-peak hours). When the environment of the monitoring area changes, the normal parameter range can be updated with appropriate traffic parameters, so as to maintain the normal parameter range in the current state of the monitoring area. In addition, when the traffic condition sensor is initially set up, the monitoring program can be executed for the monitoring area. In addition, when it is determined that there is an abnormal condition in the monitoring area, the monitoring personnel can be notified in time to implement contingency measures, and at the same time, inaccurate monitoring results can be avoided by artificially determining whether the traffic parameters are abnormal.

Although the present invention is disclosed in the foregoing embodiments, it is not intended to limit the present invention. Changes and modifications made without departing from the spirit and scope of the present invention belong to the scope of patent protection of the present invention. For the protection scope defined by the present invention, please refer to the attached patent application scope.

ND: Normal Distribution Model CI1, CI2, CI3: confidence intervals AVG: Average S: standard deviation T1, T2: time period

FIG. 1 is a flowchart of a traffic condition detection method according to an embodiment of the present invention. FIG. 2 is a flowchart of a method for obtaining a plurality of traffic parameters according to an embodiment of the present invention. FIG. 3A is a flowchart of a method for obtaining a normal parameter range according to an embodiment of the present invention. FIG. 3B is an exemplary diagram according to the normal distribution model shown in FIG. 3A . 4A is a flowchart of a method for obtaining a normal parameter range according to another embodiment of the present invention. FIG. 4B is an exemplary diagram according to the distribution state shown in FIG. 4A . 5A and 5B are flowcharts of monitoring procedures according to one or more embodiments of the present invention.

S10, S20, S30:

Claims (8)

A traffic condition detection method, comprising: obtaining a plurality of traffic parameters associated with a monitoring area, and obtaining a normal parameter range based on the traffic parameters, wherein at least half of the traffic parameters fall within the normal parameter range, There is a buffer space outside the normal parameter range, and the buffer space is adjacent to a boundary value of the normal parameter range; and a monitoring program is executed for the monitoring area, wherein the monitoring program includes: judging a real-time traffic parameter in the monitoring area Whether it falls within the normal parameter range; when the real-time traffic parameter does not fall within the normal parameter range, output a traffic abnormality notification associated with the monitoring area; and when the real-time traffic parameter falls within the normal parameter range , update the normal parameter range with the real-time traffic parameter, wherein when the real-time traffic parameter does not fall within the normal parameter range, the monitoring program further comprises: judging whether the real-time traffic parameter falls within the buffer zone; and when judging whether the real-time traffic parameter falls within the buffer zone; When the real-time traffic parameter falls within the buffer time, the normal parameter range is updated with the real-time traffic parameter. The detection method according to claim 1, wherein after determining that the real-time traffic parameter falls within the normal parameter range, and before updating the normal parameter range with the real-time traffic parameter, the monitoring program further comprises: the real-time traffic parameter The parameter is multiplied by a weight value greater than 1. The detection method according to claim 1, wherein the monitoring area is a first monitoring area, and obtaining the traffic parameters associated with the monitoring area includes: calculating the traffic parameters associated with a first traffic object in the first monitoring area a degree of difference between a first traffic parameter and a second traffic parameter associated with a second traffic object in a second monitoring area, wherein the type of the first traffic object is similar to the type of the second traffic object, The first traffic parameter and the second traffic parameter are of the same type; judging whether the degree of difference is not greater than a threshold; and when judging that the degree of difference is not greater than the threshold, corresponding to the second monitoring area accumulated A plurality of the second traffic parameters are used as the traffic parameters. The detection method according to claim 1, wherein obtaining the normal parameter range based on the traffic parameters comprises: forming a normal distribution model with the traffic parameters; and using a confidence interval of the normal distribution model as the normal parameter scope. The detection method according to claim 1, wherein obtaining the normal parameter range based on the traffic parameters comprises: filtering out a plurality of time period traffic parameters associated with the time period separation parameter from the traffic parameters according to a time period separation parameter; A normal distribution model is formed with the traffic parameters of the time period; and a confidence interval of the normal distribution model is used as the normal parameter range. The detection method according to claim 1, wherein when obtaining the traffic parameters, the detection method further comprises: recording a time parameter corresponding to each of the traffic parameters, and obtaining the normal parameter range based on the traffic parameters including: filtering out a plurality of time-period traffic parameters associated with a time period from the traffic parameters according to a distribution state of the traffic parameters corresponding to the time-parameters; forming a normal distribution model with the time-period traffic parameters; and using the time-period traffic parameters A confidence interval of the normal distribution model is used as the normal parameter range. The detection method of claim 1, wherein the traffic abnormality notification includes a position information of the monitoring area. The detection method of claim 1, wherein updating the normal parameter range with the real-time traffic parameter is: updating the normal parameter range based on the real-time traffic parameter by Bayesian inference.

Images